Release 4.0.0

MANIFEST.in

@@ -4,3 +4,5 @@ include requirements.txt
 include test.py
 recursive-include docs *
 recursive-include tests *
+global-exclude *.pyc
+global-exclude __pycache__

chardet/__init__.py

@@ -16,11 +16,14 @@
 ######################### END LICENSE BLOCK #########################
 
 
-from .compat import PY2, PY3
 from .universaldetector import UniversalDetector
+from .enums import InputState
 from .version import __version__, VERSION
 
 
+__all__ = ['UniversalDetector', 'detect', 'detect_all', '__version__', 'VERSION']
+
+
 def detect(byte_str):
     """
     Detect the encoding of the given byte string.
@@ -37,3 +40,43 @@ def detect(byte_str):
     detector = UniversalDetector()
     detector.feed(byte_str)
     return detector.close()
+
+
+def detect_all(byte_str):
+    """
+    Detect all the possible encodings of the given byte string.
+
+    :param byte_str:     The byte sequence to examine.
+    :type byte_str:      ``bytes`` or ``bytearray``
+    """
+    if not isinstance(byte_str, bytearray):
+        if not isinstance(byte_str, bytes):
+            raise TypeError('Expected object of type bytes or bytearray, got: '
+                            '{0}'.format(type(byte_str)))
+        else:
+            byte_str = bytearray(byte_str)
+
+    detector = UniversalDetector()
+    detector.feed(byte_str)
+    detector.close()
+
+    if detector._input_state == InputState.HIGH_BYTE:
+        results = []
+        for prober in detector._charset_probers:
+            if prober.get_confidence() > detector.MINIMUM_THRESHOLD:
+                charset_name = prober.charset_name
+                lower_charset_name = prober.charset_name.lower()
+                # Use Windows encoding name instead of ISO-8859 if we saw any
+                # extra Windows-specific bytes
+                if lower_charset_name.startswith('iso-8859'):
+                    if detector._has_win_bytes:
+                        charset_name = detector.ISO_WIN_MAP.get(lower_charset_name,
+                                                            charset_name)
+                results.append({
+                    'encoding': charset_name,
+                    'confidence': prober.get_confidence()
+                })
+        if len(results) > 0:
+            return sorted(results, key=lambda result: -result['confidence'])
+
+    return [detector.result]

chardet/compat.py

@@ -25,10 +25,12 @@
 if sys.version_info < (3, 0):
     PY2 = True
     PY3 = False
-    base_str = (str, unicode)
+    string_types = (str, unicode)
     text_type = unicode
+    iteritems = dict.iteritems
 else:
     PY2 = False
     PY3 = True
-    base_str = (bytes, str)
+    string_types = (bytes, str)
     text_type = str
+    iteritems = dict.items